Blindsolving Discussion and Help

[LordNikolai]

UBR-UFL-RUF is an A9. There are only 2 faces you can turn for the setup, the F and B faces. Looking at half turns, F2 won't make RUF interchangeable with any any other, but B2 will make UBR interchangeable with UFL, so your setup will be on the B face, so you're left with B or B' as possible setups. B won't make UBR interchangeable with RUF so your setup is B'.

Yes, It works! Is so difficult but not impossible.
During BLD solve is better a fast solve or try to do all case optimaly but do (for now) a slow solve?
I know do a A9 is faster than do setup-pure-antisetup because It's a move plus but thinking take me more time.

It was Marcell Endrey's 19/19 former MultiBLD WR at WC11. Well, I don't know if there's a video, but I watched parts it on the livestream.

I have a steackerless zanchi I can't use It in official speed solve but I can use It in BLD solve?

 

[JasonK]

I have a steackerless zanchi I can't use It in official speed solve but I can use It in BLD solve?

Yep. Stickerless cubes are banned in speedsolving but allowed in BLD.

 

[Ickenicke]

OK. I have learned Classic Pochmann by watching Zane's tutorial.
I now want to start using TuRBo for edges. I think I can learn it pretty fast after watching some videos and other tutorials.
The only thing I can't understand is how to solve the parity after that I have been using classic pochmann for corners.
UL and UB edges are swapped, so how to do?

 

[LordNikolai]

OK. I have learned Classic Pochmann by watching Zane's tutorial.
UL and UB edges are swapped, so how to do?

If you solve corner first you can leave the last corner unsolved and so solve the edge. In the last edge cycle you have to solve the last edge you memorized and UL (if your buffer is UB). In this mode when you solve che last edge with OP (Y or J perm) UB and UL are swapped and you fix the parity.
Excuse me if my explanation or my english isn't good, it isn't my native language.

 

[Ickenicke]

If you solve corner first you can leave the last corner unsolved and so solve the edge. In the last edge cycle you have to solve the last edge you memorized and UL (if your buffer is UB). In this mode when you solve che last edge with OP (Y or J perm) UB and UL are swapped and you fix the parity.
Excuse me if my explanation or my english isn't good, it isn't my native language.

I understand, thank you!

 

[Rpotts]

I'd transition to M2 edges and just keep with OP corners for now, it can be super fast, Cornelius Dieckmann can get ~45 seconds using it.

Make sure to also start using J/L perms to shoot to UFL and UBR, fewer moves and generally quicker than setup-Y perm-setdown. Also, anytime you have 2 consecutive targets on the U face you can use an A perm to solve both of them. Little things like that can make OP corners so much faster, and M2 will definitely save you time compared to OP edges, much fewer moves.

 

[Zane_C]

I'm getting really comfortable with the classic Pochmann method taught by Zane. 3BLD is becoming my main event now, I haven't even practiced any other event since I did my first success.

Haha, it can get quite addictive, just like any other event/puzzle.

Now my question is what method should I learn next? I really like the idea behind TuRBo but then I don't know if it's faster than M2.

I think with practice, TuRBo and M2 can be equally as fast as each other. However, M2 seems to be easier to 'master', since the setups can be simpler than those used in TuRBo.

A couple of points:
-M2 obviously requires you do be good with M-slice moves.
-TuRBo also uses M-slice moves, but is more heavy on outer-layer turns.

You might also want to think ahead and consider the possibility that you may one day switch to speed-optimal 3-cycles (3style):
-If you use UF as the buffer, you can take advantage of many R U (or L U if you prefer) 3-cycle algs.
-If you use DF as the buffer, you can take advantage of many M-slice algs available.

The choice is yours. Just remember, it's not like this decision has to be permanent. If you try TuRBo/M2 and don't like it that much, it's not a big deal. The result will be knowing a new method, possibly some additional 3-cycle algs and you'll be more skilled at applying setup moves to certan situations.

For corners, I'm thinking either R2 or maybe TuRBo for corners as well? If you have any other suggestion than the ones I mentioned, please tell me. Thanks.

It might be easier if you first focus on your new edge method, then later change to a new corner method. If you're fine with learning 2 methods at once, then do it.

Here's my response to a similar question relating to just TuRBo corners.

TuRBo corners can be very fast if you make an improvement to the standard method. Rather than limiting yourself to setting up the targets at UFL/UBR corners, you can learn how to solve any 3-cycle as long the 3 corners are on the same layer. To my understanding Kai Jiptner does this.

In my opinion, if you really want to learn TuRBo corners, once you know the 18 cases you should certainly try advancing the method as I described. I'm not a fan of standard TuRBo corners since I think it requires more effort to learn than commutators. R2 is a much simpler than comms or TuRBo, so perhaps you can use R2 as a 'stepping stone'.

 

[LordNikolai]

Other 3 question for you:
1) I'm passing to solve edges from M2 to commutator. After it I solve 3BLD in 1:25 (time memorization was 25-30) now 2:30(with the same memorization). Will I take a lot of time before I return to do sub 1:30 solve?
2) What's the difference between speed-optimal 3-cycles, BH and commutators?
3) Up to what is possible to reduce solution time with commutators?
Thanks for answers

 

[AbstractAlg]

Now my question is what method should I learn next? I really like the idea behind TuRBo but then I don't know if it's faster than M2. For corners, I'm thinking either R2 or maybe TuRBo for corners as well? If you have any other suggestion than the ones I mentioned, please tell me. Thanks.

I would suggest learning M2, and after that if you wish learn TuRBo edges, but M2 should be fast enough to reach sub45.

For corners, I started doing A-perms with setup moves. It's like 3OP but without orientation part and it turns to be extremely fast in my case. After practicing that a bit you can easily switch to TuRBo corners and after that to BH. I wouldn't suggest R2, it's in my opinion, useless.

You may want to try this one: http://www.speedsolving.com/forum/showthread.php?9594-Freestyle-Corners-for-Dummies-R2-Variant

The choice is yours, and no-one knows what is best for you unless you try it on your own.

Good luck.

 

[Zane_C]

Other 3 question for you:
1) I'm passing to solve edges from M2 to commutator. After it I solve 3BLD in 1:25 (time memorization was 25-30) now 2:30(with the same memorization). Will I take a lot of time before I return to do sub 1:30 solve?

The rate of your improvement may vary depending on a few factors:
-your ability to recognise commutators
-the move count of your commutators
-how finger trick friendly the commutators are - (M2 is quite finger trick friendly, where as some commutators are not. Thus, M2 might be faster in certain situations.)

2) What's the difference between speed-optimal 3-cycles, BH and commutators?

BH: Commutators that solve 3-cycles move-optimally.
Commutators: Any algorithm in the form of ABA'B', regardless of move count.
Speed-optimal 3-cycles: This term is used for 3-cycle algorithms that are both efficient in move count and finger trick friendly. Speed-optimal 3-cycles aren't always move-optimal and they often aren't commutators. (ie. The popular 2 gen U-perm is not optimal nor a commutator, but due to it's fast execution it is considered a speed-optimal 3-cycle)

Note: Some people (pretty much Aron and myself ) abbreviate 'speed-optimal 3-cycles' as '3style'.

3) Up to what is possible to reduce solution time with commutators?

Much like the first question, I can't really answer this because the potential reduction in time depends on the solver:
* If you become very fluent with commutators and are able to make them finger-trick friendly and relatively optimal, the improvement from M2 could be quite dramatic (25-40% faster).
* But if you remain slow and use commutators that are NOT finger-trick friendly and are far from optimal, the improvement might be very little (or not an improvement at all).

Don't worry about the last point I made, it's expected that you'll be slow at first. Over time you will gradually discover faster alternative commutators, and nice rotations to enhance the execution.

 

[LordNikolai]

-your ability to recognise commutators

I can develop it with practice.

-the move count of your commutators

I try to do all cases optimal

-how finger trick friendly the commutators are - (M2 is quite finger trick friendly, where as some commutators are not. Thus, M2 might be faster in certain situations.)

Speed-optimal 3-cycles: This term is used for 3-cycle algorithms that are both efficient in move count and finger trick friendly. Speed-optimal 3-cycles aren't always move-optimal and they often aren't commutators. (ie. The popular 2 gen U-perm is not optimal nor a commutator, but due to it's fast execution it is considered a speed-optimal 3-cycle)

Much like the first question, I can't really answer this because the potential reduction in time depends on the solver:
* If you become very fluent with commutators and are able to make them finger-trick friendly and relatively optimal, the improvement from M2 could be quite dramatic (25-40% faster).
* But if you remain slow and use commutators that are NOT finger-trick friendly and are far from optimal, the improvement might be very little (or not an improvement at all).

Don't worry about the last point I made, it's expected that you'll be slow at first. Over time you will gradually discover faster alternative commutators, and nice rotations to enhance the execution.

Thanks for yours excellent answers
Do you tell me I have to do some rotation to have a best fingertick commutator? (for example to move E or S I can rotate and use M)
Yes it works! :tu

Finally, if I want to use 3-style I have to use optimal 3-cycle but if there is an other mode not optimal and it's better I use it, than I have to solve commutators using the best mode for fingertick. I want to do learn it immediately!

I love BLD

An other fast question: How much time do you spend to have costant sub1 solves using 3-style?

 

[Zane_C]

Thanks for yours excellent answers
Do you tell me I have to do some rotation to have a best fingertick commutator? (for example to move E or S I can rotate and use M)

No problem!
Yes, that's exactly right.

I want to do learn it immediately!

Great! If you have difficulty with some cases, riffz has a list of DF 3style algs/comms here.

An other fast question: How much time do you spend to have costant sub1 solves using 3-style?

I don't cube as much as I use do, but every week I still compete in the BLD race and weekly comp, so that's at least 23 BLD attempts per week. I think I am capable of doing a lot less and still remain sub-1, but I've never done any less so I don't know for sure.

 

[Mike Hughey]

I have a separate set of images for single letters. It's just 24 letters, so it's not that many extra images. My single-letter images are all "persons", and I have four groups of them, that I use in different cases (depending on the type of piece for big cubes, depending on the cube number for multiBLD). My four groups of persons are: cartoon characters, people I know, musicians, actors/actresses.

So I never have trouble confusing a two-letter image with a single letter. And I think that's very helpful, so I recommend having at least one set of single-letter images. If you are going to the trouble to have all possible two-letter images, it's really not much more to add a set of single-letter images to the overall list.

 

[Zane_C]

In the middle of writing this response, my hand somehow slipped and I clicked 'reply to thread', clearing everything... all well.

Question to those who use letter pairs. What do you do when you have a single letter at the end?

I use another letter from the same piece to form a letter-pair. I'll set up an example:

UL=A, UB=B and LU=C
The cycle is UL>UB>LU, in other words A>B>C.

The way I would memorise this is: AB CA.
Because C (LU) is the last target, and the sticker A (UL) is on the same piece.

For this to work, you'll have stop using the beginner method I showed for forming pairs.

Also, when creating a letter pair list, should I create two lists?

Having a separate list for audio pairs like Chris has done could help. However, sounds/words can become very easy to make on the spot without the assistance of a list. Unless you're intending on using proper words and perhaps forming coherent sentences, I don't think it's necessary to form a list of audio pairs.

In the December of 2010, I was experimenting with audio/sentences and looked forward to having a monosyllable word (2 syllables at the absolute most) for every letter pair. I used Chris' list as a template, then replaced any words or sounds that didn't suit me. It took a lot of effort and I was wasting my time, I soon dropped the whole idea of having an audio list.

There are 2 ways I now form audio pairs.
- If the letters are 2 consonants; place a vowel in-between as a filler. (Example: TN = tin)
- If there is a vowel involved; say the letter pair how it would sound, as if it was a word. (Example: BU = Boo)

Note: For audio pairs beginning with 'X', it might be useful if you memorise sounds/words for that.

I guess many of the images would translate over to the audio if they have one syllable, or would that also cause confusion if BU (bull) is used for both lists?

That shouldn't cause any confusion, because it's easy to distinguish the difference between imagining a 'bull' and saying the word "bull".

To clarify: if I get bull in my memo, while I'm executing, how would I know if it's target B or BU? Or even worse when you get BU for corners as well and you have Bull two or maybe even 3 times in your memo.

If you want to use the method I described, you won't have this problem and the maximum recurrence of the same word will be 2 (once during edge memo and once during corners).

Does that ever cause confusion or does it just come with practice. Sorry, just want to clarify everything before I make a list of 528 + 504 words, thanks.

Repetition of pairs isn't a problem and your questions are good ones. :tu

 

[TMOY]

The popular 2 gen U-perm is not optimal nor a commutator

Actually it is one ([R U R U R, y2 z] conjugated by R2 U). Ok nobody will ever execute it that way.

 

[AbstractAlg]

where to learn 3-style corners?

 

[Penguino138]

I was trying to get my first successfull BLD solve using M2 for the edges and old pochman (y perm) for corners. I thought I was done so I opened my eyes. But I came to see that FD and UB were swapped, as well as URB and FUB. (I might not be using the correct sequence of letters for these pieces. I use a different memo then most) So my question is what happened? I had parity after corners so I did my R perm. But I keep coming across this case. I thought you just used M2 to switch them but in the end it does nothing. Hopefully someone can help. If you need further explanation I can do that.

 

[aronpm]

I was trying to get my first successfull BLD solve using M2 for the edges and old pochman (y perm) for corners. I thought I was done so I opened my eyes. But I came to see that FD and UB were swapped, as well as URB and FUB. (I might not be using the correct sequence of letters for these pieces. I use a different memo then most) So my question is what happened? I had parity after corners so I did my R perm. But I keep coming across this case. I thought you just used M2 to switch them but in the end it does nothing. Hopefully someone can help. If you need further explanation I can do that.

If you did an R perm to fix parity, and you ended up with 2 edges and 2 corners swapped at the end (I don't understand what corners you are talking about; there is no FUB corner, because there is no corner with stickers on both the F and B faces...), then you didn't have parity at the start.

 

[Penguino138]

If you did an R perm to fix parity, and you ended up with 2 edges and 2 corners swapped at the end (I don't understand what corners you are talking about; there is no FUB corner, because there is no corner with stickers on both the F and B faces...), then you didn't have parity at the start.

I hold my cube with yellow on top, green in front. The two edges that were swapped were the White/Green and the Yellow/Blue. The two coreners that were swapped were the Yellow/Green/Orange and the Yellow/Blue/Orange. Does that help?

 

[Litz]

Assuming you have parity, when you finish the corners and do a R-perm, those two corners will be swapped. If you solve all the edges correctly, solve edge parity and undo the R-perm at the end the cube will be solved. It looks to me like you messed up something in the edges and forgot to undo the R-perm at the end but it could have been something else.

Since you say you keep coming across that case, whenever you solve BLD, write the scramble down and memo the cube. After you memo, write the memo down and then start solving. If it happens again, follow the scramble/memo you wrote down to see what you did wrong.